On Mon, 16 Apr 2018, Mikulas Patocka wrote: > If you boot with slub_max_order=10, the kmalloc-8192 cache has 64 pages. > So yes, it increases the order of all slab caches (although not up to > 4MB). Hmmm... Ok. There is another setting slub_min_objects that controls how many objects to fit into a slab page. We could change the allocation scheme so that it finds the mininum order with the minimum waste. Allocator performance will drop though since fewer object are then in a slab page.
